ESTRO 2020 Abstract book

S67 ESTRO 2020

diagnosis, androgen deprivation therapy and risk group based on GS, T stage and PSA level, RTL quartiles were still a significant predictor of higher overall mortality (HR 1.22; 95% Cl 1.07 – 1.39; p = 0.003). Multivariate Cox regression analysis also showed a significant association of RTL quartiles with cancer specific mortality (HR 1.27; 95% CI 1.00 - 1.60; p = 0.047). Conclusion Longer leucocyte RTL correlates with higher mortality in prostate cancer patients treated with radiotherapy.

Poster Highlights: Poster highlights 4 PH: Intrafractional motion

PH-0121 Selecting the optimal fiducial marker to reduce the PTV margins for partial breast irradiation N. Hoekstra 1 , S. Habraken 1 , A. Swaak - Kragten 1 , M. Hoogeman 1 , J. Pignol 2 1 Erasmus MC Cancer Institute, Radiation Oncology, Rotterdam, The Netherlands ; 2 Dalhousie University, Radiation Oncology, Halifax, Canada Purpose or Objective In adjuvant partial breast irradiation (PBI), the geometric accuracy of the treatment is critical to avoid geographical miss. This is even more important if the number of fractions is reduced. The tumor bed is often not clearly visible on pretreatment imaging, so fiducials can be used for daily image-guidance. There are various types of fiducials available, and it is currently unknown what the impact is of a given fiducial type on the PTV margin. The purpose of this study is to select the optimal fiducial for patient positioning in PBI regarding the required PTV margin. Material and Methods Fourteen patients from the CK-APBI trial (NL6802) were included and ten patients analyzed, excluding the first 4 participants assuming a learning curve. All patients had ≥3 tantalum surgical clips placed during lumpectomy, 3 interstitial gold markers placed outside the lumpectomy area, and a gold marker taped on the areola skin. Breath- hold planning CT and 5 daily CT scans were available for all patients. The daily CT scans were acquired with an in- room CT-on-rails integrated with a CyberKnife system. The surgical clips, interstitial markers, and skin markers were identified in all CT scans. In MIM (version 6.9.3), the tumor bed in the planning CT was registered to the tumor bed in each daily CT. Subsequently, the displacement of the center of mass (CoM) per set of fiducials was calculated. This gives the residual error for the use of fiducials for daily image guidance. We calculated the group mean residual error M , the standard deviation of the systematic error Σ , the standard deviation of the random error σ , and the required margin according to van Herk (2000) for a 5- fraction and 15-fraction schedule. Results The displacements of the CoM of the different fiducial types with respect to the tumor bed are shown in Table 1. The group mean error was not significantly different from zero. The skin marker showed the largest systematic and random errors, followed by the interstitial markers. Figure 1 shows the margins required to correct for fiducial displacement for a 5-fraction and 15-fraction schedule. The margins required were ≤ 2 mm in all directions for the surgical clips. Interstitial markers required a larger and anisotropic margin, of up to 5 mm in the left-right direction. Due to the relatively large random error, fractionation has the biggest impact on the margin for a skin marker. The average PTV volume (5 fractions) would be 113 cc for the surgical clips, 134 cc for the interstitial markers, and 169 cc for the skin marker, which is a 50% larger volume.

Conclusion Based on the data taken from in-room acquired daily CT scans, surgical clips most accurately represent the position of the tumor bed. A larger margin is required if interstitial fiducials are used and a single skin marker is insufficient to accurately localize the tumor bed for daily image guidance. PH-0122 Clinical implementation of model-based CT D. Low 1 , M. Lauria 1 , B. Stiehl 1 , A. Santhanam 1 , P. Lee 2 , A. Raldow 1 , D. O'Connell 1 1 UCLA Medical Center, Department of Medical Physics, Los Angeles, USA ; 2 M.D. Anderson Cancer Center, Radiation Oncology, Houston, USA Purpose or Objective To present the first clinical implementation of a model- based CT workflow for lung cancer radiation therapy. Material and Methods This work was motivated by the lack of quantitation and the sorting-induced artifacts of commercial 4DCT. The model-based CT workflow starts with the acquisition of 25 low-mAs fast helical CT scans (the first termed the reference image) using a 64-slice CT scanner with simultaneous abdomen-based breathing surrogate measurement. The breathing surrogate amplitude is assigned to each CT slice. Deformable image registration is used to determine the voxel-voxel motion and a breathing motion model fit to each voxel. For our system, the model uses breathing amplitude and rate as the time- dependent variables, hence is termed 5DCT. The model is subsequently used to deform the reference image to a user-selected breathing amplitude. The model residuals are used to describe overall process quality, while the original CT scans are also reconstructed to describe overall process accuracy. For the clinic, we replace the 8 phase- based CT scans with model-built scans at 8 amplitudes corresponding to 8 breathing amplitude percentiles. We evaluated the first 13 clinical patients to determine the impact and quality of the new workflow, including analyzing breathing irregularity and the model residuals. Results